A trained hawk carried sensors through a swarm of hundreds of thousands of bats, recording data that helped researchers figure out how the flying mammals manage to pick out their own high-pitched echoes from those of other chatty bats.
If you have ever carried on a conversation in a noisy crowd, you have solved what is known as the cocktail party problem, focusing on a single voice among many. This isn鈥檛 just a human issue. For bats, which hunt and navigate by emitting ultrasonic calls and measuring the difference between what they sent out and what returns, detecting their own high-pitched echoes is a matter of survival.
Bats commonly live and hunt in groups of thousands. 鈥淭hey come out in swarms, and it鈥檚 chaos,鈥� says at the University of New Hampshire. So how does each animal recognise its own echo in a crowd?
Advertisement
One long-standing theory is a 鈥渏amming avoidance response鈥�, where an individual shifts its call to minimise interference from its neighbours. 鈥淲hile there is evidence in support of this response in bats, the adjustments in call features are variable,鈥� says at Johns Hopkins University in Maryland. Some researchers argue that call adjustments may instead be a more general reaction to other bats in the environment.
Some evidence for the jamming avoidance response comes from field recordings, as well as laboratory studies where researchers electronically manipulate echoes to determine what cues bats attend to. These experiments suggest that individuals shift their call frequencies or timing to avoid overlap with others. However, the studies involved tens of bats. 鈥淲hen there are thousands, that strategy breaks down,鈥� says Kloepper.
To find out how bats solve the cocktail party problem, she needed to observe a dense swarm, where obtaining audio-visual information poses a challenge. Stationary sensors provide only a limited view, and fast-flying drones are noisy and threaten bats鈥� delicate winged bodies. So Kloepper contacted a licensed falconer to propose a bird-borne platform. A year later, the team had trained a captive-bred Harris鈥檚 hawk to fly through a swarm of bats while wearing a backpack equipped with tiny microphones and video cameras.
Guided by humans positioned along its path, the hawk flew 23 passes through 600,000 bats that emerged from Jornada caves, New Mexico, for their evening meal. Its backpack recorded the bats as they travelled down the middle of a canyon.
From these recordings, the researchers extracted characteristics of each bat鈥檚 鈥渧oice鈥�. They teased out extremely small differences from one individual鈥檚 call to the next and found that these tiny variations can drive enormous differences in the signal that returns to a bat. This makes it easier for each bat to compare a returning echo to the one it sent out and ignore the input that doesn鈥檛 match. 鈥淩eally subtle changes have a big impact on whether a bat accepts or rejects the return signal,鈥� says Kloepper.
at the Max Planck Institute of Animal Behavior in Germany says it is a 鈥渧ery neat way鈥� to record the soundscape in the middle of a dense swarm. Unlike capturing individual bats and tagging them with recorders, having 鈥渁nother animal trained to fly in various paths gives you far more control on how you want to sample the soundscape鈥�, he says.
鈥淚t鈥檚 very novel to put the hawk in the mix,鈥� says Moss. 鈥淥ne question I have is whether the hawk affected the bats 鈥� if a predator is nearby, there may be effects that we can鈥檛 really detect.鈥�
Kloepper notes that a gliding raptor offers silent and mobile biomonitoring, but acknowledges that responsible falconry requires more resources than simply powering up electronic equipment. 鈥淚t鈥檚 an alternative when drones are difficult to operate due to noise levels or the surrounding environment,鈥� she says.
The work was presented at the 186th meeting of the Acoustical Society of America in Ottawa last month.
